How to create a random gamma distributed vector with a mean of 0.1 and entries between 0 and 1?

Question

Anna Igbeta 2017 年 9 月 11 日

0
リンク

この質問への直接リンク

https://jp.mathworks.com/matlabcentral/answers/356134-how-to-create-a-random-gamma-distributed-vector-with-a-mean-of-0-1-and-entries-between-0-and-1

コメント済み: Sky Sartorius 2017 年 9 月 11 日

I would like to create a random gamma distributed vector with a specific mean (0.1) and only values between 0 and 1. How ? Just how?

1 件のコメント
-1 件の古いコメントを表示-1 件の古いコメントを非表示

Sky Sartorius 2017 年 9 月 11 日

The gamma distribution is unbounded, which makes it a difficult candidate for what you're trying to do without adding steps such as clipping.

Check out the beta distribution to see if that will work for you. You can use betaincinv to help generate the vector.

サインインしてコメントする。

サインインしてこの質問に回答する。

Answer 1

José-Luis 2017 年 9 月 11 日

1
リンク

この回答への直接リンク

https://jp.mathworks.com/matlabcentral/answers/356134-how-to-create-a-random-gamma-distributed-vector-with-a-mean-of-0-1-and-entries-between-0-and-1#answer_281136

編集済み: José-Luis 2017 年 9 月 11 日

The gamma distribution takes two parameters: the shape and the scale. The mean of the distribution will be shape/scale.

That means that for a given mean (pun not intended), there are infinitely many gamma distributions. You'd need to pick either the scale or the shape and then get the other one, given your mean.

Also, the gamma distribution is asymptotic, if you limit it to the 0-1 interval then what you have is a truncated distribution that might be a very bad approximation of the gamma distribution depending on your selected shape and scale parameters.